gptkbp:instanceOf
|
theoretical agent
|
gptkbp:basedOn
|
gptkb:reinforcement_learning
gptkb:Solomonoff_induction
|
gptkbp:complexity
|
incomputable
|
gptkbp:decision
|
maximizes expected future reward
|
gptkbp:describedBy
|
gptkb:Marcus_Hutter
|
gptkbp:field
|
gptkb:artificial_intelligence
algorithmic information theory
|
gptkbp:formedBy
|
mathematical framework
|
gptkbp:goal
|
maximizing expected reward
|
gptkbp:hasApproximation
|
gptkb:AIXItl
gptkb:MC-AIXI
|
gptkbp:hasComponent
|
agent-environment interaction
|
https://www.w3.org/2000/01/rdf-schema#label
|
AIXI
|
gptkbp:influenced
|
theoretical AGI research
|
gptkbp:introducedIn
|
2000
|
gptkbp:limitation
|
not computable in practice
|
gptkbp:property
|
incomputable
mathematically optimal
|
gptkbp:publishedIn
|
gptkb:Journal_of_Artificial_Intelligence_Research
|
gptkbp:relatedTo
|
gptkb:AGI
gptkb:AIXItl
universal artificial intelligence
|
gptkbp:rewardFunction
|
environment-defined
|
gptkbp:uses
|
gptkb:universal_Turing_machine
Bayesian updating
expectimax search
|
gptkbp:bfsParent
|
gptkb:Solomonoff's_theory_of_inductive_inference
|
gptkbp:bfsLayer
|
6
|